In a recent multicentre study, the performance of the CP3000 coagulation system was assessed for prothrombin time (PT) and international normalised ratio (INR) in comparison with that of Instrumentation Laboratories' ACL TOP-300, the Cobas t411 by Roche, the Siemens BCS, Siemens Sysmex CA-7000 and Diagnostica Stago's STA-R Evolution in routine laboratory settings.
A total of 915 samples were tested by the CP3000 system, and the results were compared using regression statistics, Pearson correlation and qualitative agreement methods.
The overall tested range was 9.9-77.1 seconds, and PT results obtained with the CP3000 system showed excellent correlation with those obtained by each comparative system: the Pearson correlation coefficient ranged from 0.96 (ACL TOP-300 and Cobas) to 0.99 seconds (Siemens BCS) (see table 1). Correlation coefficients for INR results ranged from 0.95 (Cobas) to 1.0 seconds (Siemens BCS and STA-R Evolution), as seen in Table 1 (top right).
Passing and Bablok regression analysis resulted in slopes of 1.02 (Siemens BCS) to 1.32 seconds (Cobas) for INR. The percentage of samples in the INR range of 2.0-4.0 with less than 0.5 difference in pair-wise comparison (CLSI H57-A criteria) ranged from 50% (in comparison with ACL TOP-300) to 100% (in comparison with Siemens BCS). The mean difference in the therapeutic range (2.0-3.0 INR) ranged from -0.06 (STA-R Evolution) to 0.43 (Roche t411). Formal throughput testing with routine samples containing 37% specimens with prolonged PT resulted in 255 PT/hour performances. Based on five individual tests, the average time to the first (PT) result was two minutes and 12 seconds, and for a panel of PT, activated partial thromboplastin time (APTT) and fibrinogen concentration was five minutes and 27 seconds.
INR results obtained with the CP3000 system showed high levels of agreement with those generated with the Siemens BCS, Sysmex CA-7000, STAGO STA-R Evolution, ACL TOP- 300 and Cobas t411.
The study compared PT and INR results obtained with the CP3000 Coagpia PT-N reagent to those obtained from the five aforementioned coagulation platforms. In addition, the study assessed the throughput and turnaround time of STAT results on the CP3000 system.
The 919 samples were assayed in four clinical laboratories with the CP3000 system and the comparative instruments. The following calculations were performed on this data:
Throughput was assessed for PT with normal and abnormal samples, and time to first result was measured for PT alone processed as a STAT sample, and for a panel of PT activated partial thromboplastin time (APTT) and fibrinogen assays.
Pearson correlation coefficients were equal to or greater than 0.95 for PT and INR - as shown in Table 1 - in all comparisons.
The slope and intercept of the regression line suggested the strongest concordance of CP3000 system results with those obtained by the Siemens BCS and Sysmex CA-7000 (both with Thromborel S reagent), and the least amount of concordance with those obtained by Cobas t411 and STA-R Evolution. Slope and intercept values, however, significantly improved in these two comparisons when INR was considered instead of PT, as seen in Table 1.
The results obtained from the study demonstrates that INR values generated on the CP3000 system with Coagpia PT-N reagent show strong concordance with those obtained with five other coagulation analysers. INR equivalency was successfully demonstrated in comparison with Siemens BCS and Stago STA-R Evolution. Throughput was in line with the manufacturer's claim, and the time to first results was rapid.
Predicted bias at medical decision levels was the smallest in comparison with Siemens BCS, STA-R Evolution and Sysmex CA-7000, as seen in Table 2 (middle left).
The average difference between INR values in the INR range of 2.0-3.0 was less than 0.5 in each comparison, as shown in Table 3. CLSI guideline H57-A considers two PT methods equivalent if at least 85% of the samples in the INR range of 2.0-4.0 exhibit less than 0.5 difference. This criterion was fulfilled in comparison with Siemens BCS and Stago STA-R Evolution, and was not fulfilled in comparison with Cobas t411. The number of samples in the INR range of 2.0-4.0 was two and 16 for ACL TOP-300 and Sysmex CA-7000, respectively, so the results are inconclusive, as seen in Table 3 (left).
The CP3000 coagulation system is not available in all countries, including the US.
